Modeling Of Fermentation Process Based On Metabolic Network

The components of microbial fermentation process are extremely complex.The physicochemical and biochemical reactions are mixed and overlapped.The general models of fermentation process simplify the complex reaction mechanism and lack the description of cell microcosmic information.With the development of genetic engineering and cell engineering,combined with increasingly perfect methods of genomics and metabolic network analysis methods,the fermentation process modeling methods are constantly enriched,which is conducive to further exploring the reaction mechanism of the biochemical process.On the other hand,the existing detection instruments in the fermentation process are lagging behind,and many key process variables cannot be measured online.The state estimation method based on the mathematical model can estimate the parameters that are difficult to measure in the fermentation process.And it can realize online monitoring,ensure the stable operation of the fermentation process and improve quality and efficiency.This dissertation introduces the technique of metabolic network analysis in the study of fermentation process modeling and state estimation.A mathematical model of the process is established aiming at the industrial application problems of fermentation processes such as lysine and poly-?-hydroxybutyrate.Based on the simplification of metabolic network and elementary analysis,it is established by coupling metabolic flow with kinetic model and catalytic enzyme regulation,so as to realize the state estimation of the fermentation process.The specific research contents are as follows:(1)Construct the central metabolic network and simplified metabolic network of lysine fermentation based on the metabolic analysis of lysine production.Firstly,we analyze the main metabolic pathways in cellular metabolism,glycolysis,TCA cycle,pentose phosphate,oxidative phosphorylation,and their physiological significance.Then,by combining with the technological principle,fermentation characteristics,metabolic pathways and genomics of producing lysine by corynebacterium glutamate,extract the key metabolic pathways and metabolites in the lysine metabolic network.Finally,the simplified lysine metabolic network and the corresponding stoichiometric equation are established.(2)A novel coupled modeling strategy is presented by combining the macro kinetics with the metabolic flux balance equation.The coupled model is constructed on the basis of the simplified lysine metabolic network.Firstly,we establish the balance equation of intracellular metabolites,and the macro kinetics of biomass,product and substrate are respectively represented by differential equations.Then the correlation between microcosmic metabolic flux and macrostate variables is realized by coupling the metabolic flux balance equation and macrodynamics model.The key metabolic flux specific substrate consumption rate and specific product generation rate in the coupled model are expressed by empirical equations.Finally,the model parameters are identified by intelligent genetic algorithm to find the optimal parameters.The model is in good agreement with the experimental data and provides a microcosmic perspective for the quantitative study of cell flux.(3)A state estimation method of microbial metabolic is proposed for the fermentation of poly-?-hydroxybutyrate based on hybrid cybernetic model.Firstly,the hybrid cybernetic model combines the elementary mode analysis and the cybernetic model control.It not only involves both intracellular product information and metabolic behavior,but also introduces control variables to regulate the activity and synthesis of catalytic enzymes.Based on the kinetic equation of microbial metabolism,the real-time state monitoring of microbial metabolism could implement.By means of the metabolic pathway analysis,the nonlinear dynamics of the existing hybrid cybernetic model is first linearized.Then the extended Kalman filter is used to estimate the metabolic state of fermentation process.The simulation results show that the extended Kalman filter can be used to estimate some important state variables of microbial metabolism process.